Bull. Environ. Contam. Toxicol.

(1998) 61: 182-187

© 1998 Springer-Verlag New York Inc.

Total Mercury in Muscle of Fish from Two Marshes in

Goldfields, Colombia
J. Olivero, B. Solano, I. Acosta
University of Cartagena, Environmental Chemistry Group, A.A. 6541,
Cartagena, Colombia
Received: 21 March 1998/Accepted: 16 June 1998

Among Au producing countries in Latin America, Colombia has increased the

mining of the metal exponentially in recent years. However, this mining activity is
concomitant with extensive damage to the environment. The lack of
technological processes in gold extraction leads to the unmanageable
introduction of considerable amounts of Hg into the food chain. Significant Hg
concentrations have been detected in hair samples of fishermen living in gold
mining areas (Olivero et al. 1995). Most of the Au mines in Colombia are located
in the watersheds of the Magdalena and Cauca rivers. This paper addresses the
distribution of Hg in fish from two marshes influenced by Au mining in the North
of Colombia. In addition, these data were compared to that obtained in a
relatively uncontaminated marsh downstream from the extraction areas. All
three marshes provide fishmeal not only for the fishing village population but
also for other parts of Colombia.

MATERIALS AND METHODS

The study area was three marshes located in Northwestern Colombia. The
marshes were chosen because of their different geographic characteristics in
the gold mining area, and because they are extensively used for fishing. Simiti
marsh, found in the Magdalena River watershed (7°56'-8°1' N and 73°55'-73°58'
W) and Grande marsh, situated in the Cauca River watershed (8°19'-8°22’'N
and 74°29'-74°32' W), both receive direct influence of gold mining activities, with
Grande marsh being affected the most. Capote marsh (10°18'-10°23' N and
75°5'-75°9' W), located approximately 290 km downstream from the gold mining
area served as a control. Monitoring of Hg in fish was carried out during two
different periods of 1997: April-May and August-September, with monthly mean
precipitation between 167.6-284.0 and 347.7-267.7 mm, for the first and second
period respectively (IDEAN 1997).

Fish collected for this study belonged to different trophic levels, including the
phytoplanktonic Prochilodus reticulatus magdalenae (Bocachico) and Curimata
mivartii (Vizcaina); detritivorus with tendency to zooplanktonic Triportheus
magdalenae (Arenca) and Curimata magdalenae (Pincho); the detritivorous-
carnivorous Pimelodus clarias (Barbudo), and the carnivorous Petenia
kraussii (Mojarra amarilla), Plagioscion surinamensis (Pacora), Ageneiosus
caucanus (Doncella), Sorubin lima (Blanquillo), Pseudoplatystoma fasciatum
(Bagre Pintado), and Hoplias malabaricus (Moncholo).

Correspondence to: J. Olivero

182
Fish were caught by fishermen from local areas. The body length and weight of
these fish were measured. Fish were transported to the lab, eviscerated, and
frozen at -20°C. Thawed fish were dissected with plastic knives and dorsal
muscle was used for total Hg analysis by cold vapor atomic absorption
spectroscopy after acid digestion with sulfuric and nitric acids at temperatures
between 100-110 °C (Sadiq et al. 1991) for three hours. All determinations
were made in duplicate and analytical quality was ensured by using both
certified material (Tort-1, National Research Council Canada) and recovery of
Hg in spiked samples. The analysis of Tort-1 provided a mean value of 280±20
µg/Kg (certified value 330±60 µg/Kg). The recovery for spiked samples ranged
from 84.5 to 112.5% (n=6). Mean coefficient of variation from duplicate samples
with values higher than the detection limit was 7.04±0.45%, n=151.

Data were expressed as mean ± standard error of the mean for duplicate
determinations. In samples below the detection limit (7.4 µg/Kg), a mean value
equal to half of the detection limit value was used for statistical analysis. In
order to perform multiple comparisons, normality was tested according to the
Kolmogorov-Smirnov test. Interspecies differences in tissue Hg concentration
were analyzed using one-way ANOVA and Tukey’s Test at p<0.05 significance
level.

RESULTS AND DISCUSSION

Table 1 summarizes Hg concentrations in muscles of different fish species

caught from marshes monitored during the two sampling periods.

Hg concentrations in fish from Grande marsh vary considerably with the

species; Prochilodus reticulatus magdalenae (primary consumer) contains the
lowest while Ageneiosus caucanus (tertiary consumer) the highest. For
samples collected from this marsh, Petenia kraussii Hoplias malabaricus and
Ageneiosus caucanus showed relatively higher Hg concentrations than the
internationally accepted limit of 0.5 mg Hg/Kg for fish consumption (WHO 1991).
In all samples, species belonging to different trophic levels showed significant
statistical differences (p<0.001). Significant differences between specimens
caught in different sampling periods were detected for the carnivorous
Ageneiosus caucanus.

Fish from Simiti marsh had Hg concentrations lower than that observed in those
from Grande marsh. Nevertheless, some of these fish exhibited values close to
50% the guideline limit. There were not significant differences between Hg
concentrations for the species collected in different sampling periods (p=0.228).
However, significant differences (p=0.015) were observed between the
carnivorous Blanquillo and Moncholo and the phytoplanktonic species
Bocachico during the second sampling period.

As expected, the lowest Hg mercury concentrations were found in fish from

Capote Marsh, with the carnivorous species Plagioscion surinamensis showing
the highest values. This species had Hg concentrations significantly different

183
Table 1. Hg concentrations in muscle of fish caught from marshes within gold
mining area and control marsh.

184
* Standard error of the mean. Range of data is given under the means. Number
of samples are given within brackets. DL = Detection limit: 7.4 µg/Kg.

from Prochilodus reticulatus magdalenae (p<0.001) during the two sampling

periods. As a comparison, the observed value for Plagioscion surinamensis was
similar to that detected for the primary consumer Curimata mivartii during the
first sampling period in Grande Marsh.

From Table 1 it is evident that Hg in fish from the monitored marshes increases
in the order Capote<Simiti<Grande. This spatial distribution is directly related to
the link between those waterbodies and gold mining.

It is important to mention that there are many similarities between our data and
those reported for gold mining areas in Brazil (Nriagu et al. 1992; Palheta and
Taylor 1995; Boshio and Henshel 1995; Bidone et al. 1997) especially for
mercury distribution in fish from Grande marsh. However, for this particular
waterbody, our data showed that the carnivorous Hoplias malabaricus had 50%
higher concentrations than those reported by Palheta and Taylor (1995) within
the Gurupi gold field in the Brazilian State of Para.

Despite the low Hg concentrations detected in Simiti marsh, a difference is

evident between carnivorous and phytoplanktonic species from both gold
mining-affected marshes. This suggests that a biomagnification process is
taking place in the trophic chain for those waterbodies, particularly in Grande
marsh. Figure 1 summarizes the sequential increase in Hg concentration with
the trophic position of the species. Although Triportheus maqdalenae is
considered a lower trophic level than Pimelodus clarias, it showed a high Hg

185
Figure 1. Distribution of mercury in fish from different trophic levels in Grande
Marsh. Bars on columns represent SE of the means.

concentration. This was probably due to direct bioaccumulation as observed in

low-contaminated places (Olivero et al. 1997).

Correlations between the concentration of Hg in muscle and the fish length

(r=0.995, p<0.001) and weight (r=0.984, p=0.002, n=5) were observed in
Prochilodus reticulatus magdalenae during the first sampling period in Grande
Marsh. This suggests the possibility of bioaccumulation of Hg in this primary
consumer species.

Given the fact that Prochilodus reticulatus magdalenae is the most popular and
economically important fish in Colombia and given the concentration of Hg
found in this species, consumption of this species may prove less risky.
However, the reality may be different. Frequent fish consumption may increase
the overall Hg intake to dangerous levels, particularly in fish caught from Grande
marsh. This may lead to neurotoxicity, previously observed even at low level
exposure (Cranmer et al. 1996; Lebel et al. 1996).

These results have shown that consuming fish from Grande marsh may result in
a high risk of Hg contamination for fishermen who depend only on this
ecosystem as their protein source.

186
Acknowledgments. We thank “Fondo para la Proteccion del Medio Ambiente
Jose Celestino Mutis, FEN Colombia” for financial assistance, fishermen from
the study area for providing fish samples, and also to Claudia Mendoza and
Rafael Olivero for technical assistance.

REFERENCES

Bidone E, Castilhos Z, Cid de Souza T, Lacerda L (1997) Fish contamination

and human exposure to mercury in the Tapajós River Basin, Pará State,
Amazon, Brazil: A screening approach. Bull Environ Contam Toxicol
59: 194-201
Boischio A, Henshel D (1996) Risk assessment of mercury exposure through
fish consumption by the riverside people in the Madeira basin, Amazon,
1991. Neurotoxicology 17:169-175
Cranmer M, Gilbert S, Cranmer J (1996). Neurotoxicity of mercury indicators
and effects of low-level exposure: overview. Neurotoxicology 17:9-14
IDEAM (1997) National institute of hydrology, meteorology and environmental
research. Hydrometeorological information system.
Lebel J, Mergler D, Lucotte M, Amorim M, Dolbec J, Miranda D, Arantes G,
Rheault I, Pichet P (1996) Evidence of early nervous system dysfunction in
Amazonian populations exposed to low-levels of methylmercury.
Neurotoxicology 17: 157-168
Nriagu J, Pfeiffer W, Malm O (1992). Mercury pollution in Brazil. Nature 356:189
Olivero J, Mendoza C, Mestre J (1995). Hair mercury levels in people from
the gold mining zone in Colombia. Rev Saude Publica 29:376-379
Olivero J, Navas V, Perez A, Solano B, Acosta I, Arguello E, Salas R (1997)
Mercury levels in muscle of some fish species from the Dique Channel,
Colombia. Bull Environ Contam Toxicol 58:865-870
Palheta D, Taylor A (1995) Mercury in environmental and biological samples
from a gold mining area in the Amazon region of Brazil. Sci Total Environ
168:63-69
Sadiq M, Zaidi T, Al-Mohana H (1991) Sample weight and digestion
temperature as critical factors in mercury determination in fish. Bull Environ
Contam Toxicol 47:335-341
WHO. Environmental Health Criteria 101 (1991) Methylmercury. Geneva

187